A Prospective Cohort Study Comparing AI Prediction Model With Imaging Assessment to Diagnose Lymph Node Metastasis in Cervical Cancer

A Prospective Cohort Study Comparing Artificial Intelligence Multimodal Fusion Prediction Models With Conventional Imaging Assessment for the Diagnosis of Pelvic Lymph Node Metastasis in Cervical Cancer

The goal of this prospective cohort study is to learn whether artificial intelligence multimodal fusion prediction models are effective in diagnosing pelvic lymph node metastasis in cervical cancer. The main question it aims to answer is: can artificial intelligence multimodal fusion prediction models improve the accuracy of preoperative diagnosis of pelvic lymph node metastasis in cervical cancer? The researchers compared the AI multimodal fusion prediction model with traditional imaging physician assessments to see if the prediction model could yield more accurate lymph node metastasis determinations. Participants will undergo pelvic MRI after pathologically confirming a diagnosis of cervical cancer, and the results will be used to determine pelvic lymph node metastasis status by the predictive model and the imaging physician, respectively. Subsequent pathology results after surgical lymph node clearance will be used as the gold standard to determine the accuracy of the two preoperative lymph node diagnostic modalities.

Study Overview

• Status: Not yet recruiting
• Conditions: Uterine Cervical Neoplasms
• Intervention / Treatment: Diagnostic test: AI Prediction Model; Diagnostic test: Conventional Imageing Assessment

• Study Type: Interventional
• Enrollment (Estimated): 230
• Phase: Not Applicable

Contacts and Locations

Study Locations

• China
  • Shanghai
    • Shanghai, Shanghai, China, 200090
      • The Obstetrics and Gynecology Hospital of Fudan University

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Adult; Older Adult
• Accepts Healthy Volunteers: No

Description

Inclusion criteria:

1. patients with preoperative diagnosis of invasive cervical cancer stage I-III, with any type of pathology, and patients who underwent radical/modified radical cervical cancer surgery + pelvic lymph node dissection in our hospital or sub-center;
2. Age ≥18 years and ≤80 years;
3. patients who underwent preoperative pelvic MRI (plain/enhanced) imaging in our hospital or sub-centers.

Exclusion criteria:

1. patients during pregnancy or lactation, patients with abortion within 42 days;
2. patients who are undergoing or have undergone preoperative neoadjuvant chemotherapy or radiotherapy for this cervical cancer;
3. Patients with other malignant tumors within 5 years;
4. Combination of other underlying diseases that may lead to enlarged pelvic lymph nodes;
5. patients whose preoperative pelvic MRI date is more than 1 month from the day of surgery;
6. poor quality imaging images that are unrecognizable.

Study Plan

How is the study designed?

Design Details

• Primary Purpose: Diagnostic
• Allocation: Non-Randomized
• Interventional Model: Factorial Assignment
• Masking: None (Open Label)

Number of Arms

2

Arms and Interventions

Participant Group / Arm	Intervention / Treatment
Experimental: AI Prediction Model	Diagnostic test: AI Prediction Model; Pelvic MRI was performed after pathologic diagnosis clarified the diagnosis of cervical cancer. Further pelvic lymph node metastasis status was determined by artificial intelligence multimodal fusion prediction modeling
Active Comparator: Conventional Imageing Assessment	Diagnostic test: Conventional Imageing Assessment; Pelvic MRI was performed after pathologic diagnosis clarified the diagnosis of cervical cancer.Further pelvic MRI images are read by a specialized imaging physician to determine pelvic lymph node status.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Accuracy in determining pelvic lymph node metastasis	After the subjects underwent surgical treatment, surgical pathology served as the gold standard for evaluating the accuracy of the AI predictive model in comparison to traditional imaging diagnosis. In the statistical analysis phase, sensitivity and specificity were utilized as the primary indicators to assess the accuracy of both diagnostic modalities.	The time frame was from subject enrollment until surgical pathology results were obtained. The time between subject enrollment and the availability of surgical pathology results was approximately 1 to 1.5 months.

Collaborators and Investigators

• Sponsor: Obstetrics & Gynecology Hospital of Fudan University

Study record dates

Study Major Dates

• Study Start (Estimated): August 1, 2024
• Primary Completion (Estimated): December 1, 2027
• Study Completion (Estimated): December 1, 2027

Study Registration Dates

• First Submitted: August 1, 2024
• First Submitted That Met QC Criteria: August 6, 2024
• First Posted (Actual): August 7, 2024

Study Record Updates

• Last Update Posted (Actual): August 7, 2024
• Last Update Submitted That Met QC Criteria: August 6, 2024
• Last Verified: August 1, 2024

More Information

Terms related to this study

• Additional Relevant MeSH Terms: Pathologic Processes; Neoplasms; Urogenital Neoplasms; Neoplasms by Site; Uterine Neoplasms; Genital Neoplasms, Female; Uterine Cervical Diseases; Uterine Diseases; Neoplastic Processes; Neoplasm Metastasis; Female Urogenital Diseases; Female Urogenital Diseases and Pregnancy Complications; Urogenital Diseases; Genital Diseases; Genital Diseases, Female; Uterine Cervical Neoplasms; Lymphatic Metastasis
• Other Study ID Numbers: FUOBGY-2024-64

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: NO

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No